TALENT EDUCATION

The latest research of the Zhejiang University team appeared on the cover of the Science sub-issue: aerial “swarm” robot with intelligent brain


Wen | Ma Jing, an intern of China Science Daily, and reporters Li Chenyang and Cui Xueqin

Editor| He Tao

A group of miniature drones skimmed swiftly through the low bushes, forming unhindered by branches.

When they move freely through a dense bamboo forest, sometimes flashing left and right to avoid obstacles, sometimes crossing and changing formations, like birds in the forest dancing and chasing each other.

Don’t look at these miniature drones are only the size of a palm, lighter than a cola, but they can independently plan the optimal flight route with their own visual and computing resources, while also flexibly avoiding obstacles, avoiding collisions between drones, and completing formation coordination tasks.

What’s more, when these miniature drones pass through the bamboo forest, they can draw a 3D image of the surrounding environment through the positioning algorithm. The team of the School of Control Science and Engineering of Zhejiang University and the Huzhou Research Institute are the developers of this group of miniature unmanned aerial vehicles. The study was featured on the cover of Science Robotics, an authoritative journal in the field of robotics, on May 4.

Small fuselage contains “great wisdom”

With the development of communication, sensing and other technologies, drones have long been commonplace, and large-scale UAV aerial formation performances have also become the finale of many celebration occasions.

However, the flight activities of traditional UAVs mainly rely on human control, kaleidoscopic formations, and also stem from pre-designed satellite positioning and trajectory coding. Once out of control of the ground computer, these drones will be leaderless, or even crash into obstacles.

So, is it possible to achieve autonomous intelligence of drones so that they do not have to rely heavily on human control? With this curiosity, the Zhejiang University research team found the answer from nature.

“The herringbone geese, the orderly flocks of pigeons, when we look up at the sky, the flocks of birds in nature always make us think about how they stay in formation.” Xu Chao, corresponding author of the paper and dean of the Huzhou Research Institute of Zhejiang University, said.

Birds are able to maintain relatively long-term, smooth flight movements, while having keen visual and motor perception, as well as high degrees of freedom and flexibility. Taking the flight principle of birds as a reference, combined with the “time and space optimization technology” for trajectory optimization, the Zhejiang University team spent more than two years to solve a series of core technologies such as intelligent navigation and rapid obstacle avoidance methods of robot single machines and groups in unknown and complex environments, and developed a set of small, lightweight, autonomous and controllable and group-oriented flight system – intelligent aerial robots.

The intelligent aerial robot is made up of many miniature drones, and a single drone is only the size of a palm, which is lighter than the weight of a Coke. In the case of only using onboard vision and onboard computing resources, it can independently plan the optimal flight path, while meeting the requirements of flight efficiency, avoiding obstacles, avoiding collisions between UAVs, and formation coordination.

“We make the robot understand the world, actively avoid obstacles, and achieve group flight.” Co-corresponding author Dr. Gao Fei from the School of Control Science and Engineering of Zhejiang University said.

“We’ve also built an intelligent brain for aerial robots.” Xu Chao introduced that although the processor is only the size of a thumb, it can independently calculate and process the massive amount of information encountered during flight. “The key to characterizing the motion information of the physical world into a series of mathematical problems is to identify the hidden special mathematical structure behind the complex problems, simplify the complex, and use the overstretched on-board computing resources to perfectly solve the problem.”

A “generalist” who performs a task

In science fiction movies, similar “swarm” drones are always used as a tool to attack or assassinate targets. However, reality technology emphasizes the birth of good, and the intelligent aerial robot developed by the Zhejiang University team is expected to play a role in a variety of real-world scenarios.

The most promising applications are some disaster relief scenarios. For example, in disasters such as fires, earthquakes and floods, the environment is harsh, dangerous, and difficult to provide disaster relief, at this time, the use of swarms of drones can be multi-faceted search and guidance, under the premise of ensuring the personal safety of rescuers, collect on-site information, and provide emergency supplies to trapped personnel.

Intelligent aerial robots can also replace researchers to enter natural areas and carry out ecological research. Small and flexible drones can quickly reach some areas that humans cannot tread, but also minimize the interference and damage of human activities to natural space.

In addition, scientists have a more avant-garde imagination: in the future, humans will realize the deployment of air transportation systems, with the help of this intelligent aerial robot, vehicles can flexibly travel between dense buildings and in the interweaving with other vehicles…

Xu Chao introduced that from drawing design to hardware debugging, from algorithm research and development to code writing, from experimental testing to system optimization, it is completely completely completed by the teachers and students of the Zhejiang University team, “In the world, there are only a few teams that can independently complete the whole process of work.” ”

In the process of research and development, the team has made breakthroughs in the development of intelligent, networked, autonomous and other aspects of technology, and this achievement will also have a positive effect on the machine substitution and industrial upgrading of the industrial world.

Reference Links:

https://www.science.org/doi/10.1126/scirobotics.abm5954

https://mp.weixin.qq.com/s/1gYCNzFumi-WrujCkY0kvg

https://news.sciencenet.cn/sbhtmlnews/2022/5/369272.shtm
 
Special statement: This article is reproduced only for the purpose of disseminating information, does not mean to represent the views of this website or confirm the authenticity of its content; if other media, websites or individuals reprint from this website, you must retain the “source” indicated on this website, and bear the legal responsibility such as copyright; if the author does not want to be reproduced or contact us for reprint fees, please contact us.



Source link

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button